Abstract Cancer is now the number one cause of death in this country. Many recently developed treatments have focused on managing the growth of existing tumors, and some of these treatments have turned previously lethal cancers into manageable conditions. The study of tumors has led to an in depth characterization of existing disease and the identification of many genes that are linked to cancer. The majority of what we know about cancer is based on existing tumors, but there is little knowledge on the mechanisms behind tumor initiation. While we know a great deal about the genetic hits and environmental insults that lead to tumor formation, it remains unclear which cells serve as cancer cells of origin. Can any cell make a tumor? Most adult tissues contain a wide variety of cell types including stem cells, transit-amplifying cells and differentiated cells. Many have proposed that adult stem cells are the most likely cancer cell of origin because of their longevity in the tissue and their capacity for self-renewal, or unlimited growth. Furthermore, many of the pathways thought to participate in tumorigenesis have also been shown to play a role in stem cell self-renewal in adult stem cells. However, the identity of the particular cell types targeted in tumorigenesis is obscured by the fact that most studies of tumorigenesis utilize pre-existing tumor tissue, a retrospective approach. Until recently there were no tools available to deliver genetic hits to specific cell types in a tissue to ask which could serve as cancer cells of origin. In addition, there are few model systems that allow for the isolation of cells from tissue undergoing transformation that faithfully mimic natural tumorigenesis. In short, the accumulated literature on tumorigenesis relies a great deal on data accumulated from transformed tissue targeted by an unknown mechanism in an unknown cell type. We have designed a model system that takes a prospective approach to identifying cancer cells of origin. We are taking advantage of the latest tools and genetic tricks to target genetic hits to particular cell types in the epidermis, the most common target tissue of cancer. With these tools, we are applying genetic hits to either stem cells or their transit-amplifying progeny to probe which is better able to serve as a cancer cell of origin. In addition, this model system allows for the purification of the targeted cells at any point during tumor initiation or progression. This will enable us to identify which genes are affected by tumor initiation, and which are specifically altered in benign versus malignant tumor initiation. These findings should prove critical to not only the treatment of cancer but potentially even the prevention of cancer. In addition, this work will unearth a wealth of information about stem cells and their progeny, and whether pathways that regulate stem cells are exploited by cancer.